Modeling and Simulation of Three Phase Fluidized Bed ...

International Journal of Bioscience, Biochemistry and Bioinformatics, Vol. 2, No. 6, November 2012

Modeling and Simulation of Three Phase Fluidized Bed Characteristics Using MATLAB A. Sivalingam, T. Kannadasan, M. Thirumarimurugan, and D. Prabhakaran 

Abstract—In literature, several correlations are available for the prediction of mass transfer characteristics. However, almost no correlations have been so far proposed counting the effect of Archimedes number, which is widely used in calculating the different parameters influencing mass transfer in fluidized bed. Hence an effort has been made by carrying out mass transfer studies in a co-current three phase fluidization system for Newtonian and non Newtonian fluids with and without internals, to propose correlations for Sherwood number relating with solids holdup, Reynolds number and Archimedes number. In this study, various sizes of Gypsum particles, air and water were used as solid, gas and liquid phases respectively. Water was replaced with various concentrations of Carboxymethylcellulose for non Newtonian studies. Experiments were carried out with liquid as continuous phase and gas as discrete phase. The bed height and manometer readings were noted for different gas and liquid velocities. From the recorded data, phase holdups and mass transfer characteristics were determined. From the experimental data, correlations were developed for Sherwood number relating with Archimedes number by dimensional analysis and linear regression analysis. The simulated values were compared with experimental values and verified with R2 values. Index Terms—Archimedes number, fluidized bed, gypsum particles, phase holdups.

I. INTRODUCTION Fluidization has been a pivotal topic of research among the researchers for the past three decades [1]. Since the Winkler process for gasification of coal was on track, the focus on fluidization has been consistently gaining significant importance in industrial phase as well as in academic research. The concept of fluidization in Winkler process was further developed into various successful applications such as FCC (fluid catalytic cracking), Metallurgical process, Sohio process, synthesis of polymers, combustion and in the recent years for the treatment of waste water [1]-[5]. Analogous to such industrial developments, considerable academic research has also been carried out in the field of fluidization. Relative to other multiphase contactors, there are plenty of works available in the three phase fluidization, with most studies directed towards understanding the hydrodynamic characteristics [6]-[8]. Comparatively very few researchers have worked on phase studies as well as mass transfer in such systems [9], [10]. However, significant works have been carried out by researchers so far focusing on the prediction of Manuscript received May 22, 2012; revised October 23, 2012. The authors are with the Department of Chemical Engineering, Coimbatore Institute of Technology, Coimbatore -641014, India (email: [email protected], [email protected], thirumarimurugan@ gmail.com, [email protected]).

411

Sherwood number in three phase systems by proposing different correlations [11]. In this present work, the mass transfer characteristic, Sherwood number of a co-current three phase fluidization system has been studied for different sizes of gypsum particle with both Newtonian and non Newtonian fluids. An effort has been made to study the effect of internals on the Sherwood number by carrying out experiments with and without internals. From the experimental data, correlations were developed for Sherwood number using dimensional analysis and linear regression analysis, which comprise of Archimedes number, Reynolds number and solid holdup. Archimedes number is generally used in calculations of fluid motion due to density difference. According to Goossens [12], the calculation of the Archimedes number is a relatively simple tool for industry to predict the class and the related fluidization behavior of any fluid particle system under consideration. Gupta investigated several correlations for estimation of minimum fluidization velocity, in which Archimedes number has been effectively dealt with [13]. Hence, a significant attempt was made to develop correlations for relating the mass transfer characteristic-Sherwood number with viscous forces and solid holdup for both Newtonian and non Newtonian fluids, in the presence and absence of internals. The regression analysis was carried out with the aid of MATLAB programming and the simulated values were compared with the experimental values graphically and subsequently verified with the R2 values. Notations used: NAr - Liquid Archimedes number NRe,m liq - Modified Reynolds number for liquid, NSh - Sherwood number, NSh = Kya ε dp / Dv a εs - Solid holdup ρl - Liquid density, kg/m3 ρg - Gas density, kg/m3 ρm - Density of manometric liquid, kg/m3 ρs - Density of solid particles, kg/ m3 CMC - Carboxymethylcellulose R2 - Linear Regression Coefficient

II. MATERIALS AND METHODS The experimental setup consists of a liquid storage tank from which water is pumped by a centrifugal pump. The flow rate of water and air is measured using a rotameter and an orifice meter, respectively. A schematic diagram of experimental setup is shown in Fig. 1. A vertical Perspex column, 1.6 m tall with internal diameter of 5.410-2 m and

International Journal of Bioscience, Biochemistry and Bioinformatics, Vol. 2, No. 6, November 2012

outer diameter of 610-2 m is used, which consists of three sections: the gas-liquid distributor section, test section and gas-liquid disengagement section. The gas-liquid distributor is located at the bottom of the test section and designed in such a manner that uniform distribution of the liquid and gas can be maintained in the column. The circular gas distributor section made of copper is provided with four protrusions each having I.D. of 1.4 10-2 m. The liquid inlet pipe of 2.5 10-2 m I.D. is located centrally in this section. The outlet of the test section is at a height of 1.5 m and has a mesh attached to it in order to retain the entrained particles.

TABLE I: PROPERTIES OF GYPSUM PARTICLES USED. Particle size dp(m)

True density kg/m3

0.000842

2240

0.001676 0.002818 0.0007671 0.001753 0.002591

2143 2190 2849.7 2744.6 2727.5

TABLE II: PROPERTIES OF AIR AND WATER. Fluid Air Water

Densit y Kgm-3 1.15 995.6

Viscosit y Nsm-2 0.000019 0.0085

TABLE III: PROPERTIES OF CMC USED. CMC Conc.% 0.1 0.2 0.3

Density kg/m3 1180 1200 1230

Viscosity measurement n K kg/ms2-n 0.0081 0.93 0.02 0.88 0.26 0.85

Fig. 1. Experimental setup.

There are two pressure tapings provided at the top and the bottom of the test section which are connected to the manometer for pressure drop measurement. Mercury and Carbon tetra chloride are used as manometer liquids for measuring the pressure drop across the rotameter and the orifice meter, respectively. Gypsum particles of sizes 0.084210-2 m, 0.167610-2 m and 0.281810-2 m are used for Newtonian fluid, water and 0.076710-2 m, 0.175310-2 m and 0.259110-2 m are used for 0.1%, 0.2% and 0.3% of Non Newtonian fluid, CMC for mass transfer studies. The solid particles are supported on a perforated stainless steel mesh containing 300 spaced holes of 0.0510-2 m each. The experiment is conducted with internals and without internals. Internals are coaxially placed ten spherical ceramic balls of 2.510-2 m diameter fitted on a rod which is used as a promoter. The liquid velocity is kept constant and the gas velocity is varied. For each gas velocity, the fluidized bed height and manometer readings are noted when steady state is attained. The same procedure is repeated for four different liquid velocities with and without internals. Volumetric flow rate, superficial gas velocity and the corresponding bed heights are measured. After thrice the residence time, about 60 ml of the sample solution that is coming out of the fluidized bed is collected and analyzed for its total calcium sulphate content by volumetric titration method.

Fig. 2. NSh vs NRe,m,gas (Without internals).

Fig. 3. NSh vs NRe,m,liq (Without internals).

However, the Fig. 6 and Fig. 7 show that the effect of Modified Reynolds number on the Sherwood number is less pronounced in the case of CMC solution. In the case of water, the presence of internals increases the Sherwood number but for CMC solution, the presence of internals decreases the Sherwood number. For both water and CMC solution, the Sherwood number increases with increasing particle size.

III. RESULTS AND DISCUSSION The experimental results obtained from mass transfer studies shows the variation of Sherwood number with the variation in important parameters considered in the study. From Fig. 2 to Fig. 5, it is observed that the Sherwood number increases with increase in Modified Reynolds number for both gas and liquid phases in the case of water.

Fig. 4. NSh vs NRe,m,gas (With internals).

412

International Journal of Bioscience, Biochemistry and Bioinformatics, Vol. 2, No. 6, November 2012

Fig. 5. NSh vs NRe,m,liq (With internals).

Fig. 10. Simulated NSh vs Experimental NSh (Without internals, Non Newtonian).

Fig. 6. NSh vs NRe,m,gas (Without internals).

Fig. 11. Simulated NSh vs Experimental NSh (With internals, Non Newtonian).

It has been observed using hydrodynamic studies that as the Modified Reynolds number (either for gas or liquid) increases, the solids and liquid hold-ups decreases, but the gas hold-up increases for both water and CMC solutions [14]–[17]. Also, it has been found that solids holdup are less, where as liquid and gas holdups are slightly more for CMC solutions than water. An increase in the gas holdup in the system means there are more gas bubbles leading to vigorous mixing, which can result in the reduction in boundary layer resistance and enhancement of mass transfer rate. These effects are more pronounced in CMC solution, which is denser and more viscous when compared to water. Hence, there is a pronounced increase in Sherwood number with increase in Modified Reynolds number for water, but the influence is marginal for CMC solutions. The presence of internals in the column enhances turbulence and mixing of phases leading to better mass transfer rates. However, these effects are more pronounced in water when compared to CMC solutions. Hence, the observed Sherwood numbers are higher for water with internals in the column. The fluidizing solids, because of their convective motion in the column, help in the breakup of bigger gas bubbles into smaller sizes and also promote efficient dispersion of these bubbles in the column. The combined motion of the smaller gas bubbles and solids can lead to higher turbulence and mixing, which in turn enhances the mass transfer from solid phase to liquid phase. Now, when the solid particle size increases it would be heavier and interactions between heavier solid particles lead to more break up of gas bubbles. Hence, this explains the observed trend of increasing Sherwood numbers with increasing particle sizes. Using the experimental data obtained for Sherwood number, empirical correlations involving dimensionless groups are proposed with the help of dimensional analysis. MATLAB 7.0 software has been used to arrive at the final expressions which are given from Equations (1) to (4).

Fig. 7. NSh vs NRe, m, gas (With internals).

Fig. 8. Simulated NSh vs Experimental NSh (Without internals, Newtonian).

Fig. 9. Simulated NSh vs Experimental NSh (With internals, Newtonian).

413

International Journal of Bioscience, Biochemistry and Bioinformatics, Vol. 2, No. 6, November 2012

Water – solids system without internals

[5]

NSh = 0.9298 (εs) 0.1379 (NRe,m liq)0.6030 (NAr)0.1046

(1)

[6]

Water – solids system with internals NSh = 0.1578 (εs)

0.1079

(NRe,m liq)

0.7515

[7]

(NAr)

0.2091

(2) [8]

CMC solution - solids system without internals NSh = 3.7818 (εs) 1.0665 (NRe,m liq)0.9869 (NAr)0.0599

(3)

[9]

CMC solution - solids system with internals [10]

NSh = 1.4822 (εs)

0.6993

(NRe,m liq)

0.8928

(NAr)

0.1250

(4) [11]

IV. CONCLUSION

[12]

Mass transfer studies were carried out in a co-current three phase fluidization system for Newtonian and non Newtonian fluids. The gaseous medium was air in all the cases. The Newtonian liquid and the non Newtonian liquids used were water and various concentrations of CMC respectively. Gypsum particles of various sizes were used as solid phase. Mass transfer characteristic of co-current three phase fluidization, were determined in terms of Sherwood number. It was observed from the investigation that the Sherwood number is influenced by gas as well as liquid superficial velocities, whether the liquid is water or CMC solution, whether column has internals or not and also the solid particle size. The presence of internals in the column has not significantly influenced the Sherwood number for mass transfer. This may be due to the fact that the geometry as well as the size of the internals chosen in the work has not contributed to the enhancement of turbulence and mixing of phases. From the experimental data, correlations were proposed for Sherwood number comprising of solids holdup, Reynolds number and Archimedes number using Linear Regression Analysis. The simulated values were compared with experimental values graphically for Newtonian and non Newtonian fluids, with and without internals. From Fig. 8 to Fig. 11, The R2 values were found to be closer to 0.9 for all the cases, which implies that the proposed correlations generate satisfactory results with minimal error.

[13]

[14]

[15]

[16]

[17]

A. Sivalingam was awarded the Ph.D degree in chemical engineering from Anna university of Technology, Coimbatore, in 2011. He has 28 years of teaching service with one year Industrial experience. He is currently the associate professor, in Department of Chemical Engineering, Coimbatore Institute of Technology, Coimbatore, Tamil Nadu. He has published more than eight papers in referred journals and presented nineteen papers in national or international conferences. He is the member of Board of Studies of Coimbatore Institute of Technology, Coimbatore and Member of Examiners in various Universities in India and also the member of Asia Pacific Chemical, Biological and Environmental Engineering Society (APCBEES). He is also a Life member of Indian Institute of Chemical Engineers and Indian society of Technical Education.

T. Kannadasan was awarded the Ph.D. degree in chemical engineering from Annamalai University, Chidambaram. He has served as Dean of Academic Campus, Anna University of Technology, Coimbatore; the director/academic courses, Anna University of Technology, Coimbatore; and as the Vice Chancellor (Officiate) of Anna University of Technology Coimbatore. Currently he is the professor and Head of Chemical Engineering Department at Coimbatore Institute of Technology, Coimbatore, Tamilnadu, India. He has published number of papers in reputed national and international journals. He has guided 20 M.Tech. Thesis in various areas of chemical engineering. He has guided 13 Ph.D. students out of which 8 are completed and 5 are still going on. He is a Life member of Indian Institute of Chemical Engineers (IICh. E), Indian Society for Technical Education and The Institution of Engineers (India), senior member of Asia Pacific Chemical , Biological & Environmental Engineering Society (APCBEES). He is the editor of Special Issue of “Nano Trends – A Journal of Nanotechnology and its Applications”, 2006.

REFERENCES [1]

[2]

[3]

[4]

M. S. N. Murthy, K. V .Ramesh, G. Prabhakar, and P. Venkateswarlu, “Phase holdups in three-phase fluidized beds in the presence of disc promoter,” Experimental Thermal and Fluid Science, 2010. A. Macchi, H. Bi, J. R. Grace, C. A. McKnight, and L. Hackman, “Dimensional hydrodynamic similitude in three-phase fluidized beds,” Chemical Engineering Science, vol. 56, pp. 6039-6045, 2001. A. H. Sulaymon, T. J. M. Ali, and H. Jawad, “Hydrodynamic characteristics of three-phase non newtonian liquid-gas - solid fluidized beds,” Emirates Journal for Engineering Research, vol. 15, no. 1, pp. 41-49, 2010. J. P. Zhang, N. Epstein, and J. R. Grace, “Minimum fluidization velocities for gas-liquid-solid three-phase systems,” Powder Technology, vol. 100, pp. 113-118, 1998. I. Nikov1 and H. Delmas, “Mechanism of liquid-solid mass transfer and shear stress in three-phase fluidized beds,” Chemical Engineering Science, vol. 47, no. 3, pp. 673-681, 1992. D. C. Arters and L. S. Fan, “Solid-liquid mass transfer in a gas-liquid-solid fluidized bed,” Chemical Engineering Science, vol. 41, no. 1, pp. 107-115, 1986 S. K. Chang, Y. Kang, and S. D. Kim, “Mass transfer in two- and three-phase Fluidized beds,” Journal of Chemical Engineering of Japan, vol. 19, no. 6, pp. 524-530, 1986. W. R. A. Goossens, “Classification of fluidized particles by Archimedes number,” Powder Technology, vol. 98, pp. 48-53, 1998. S. K. Gupta, V. K. Agarval, S. N. Singh, V. Seshadri, D. Mills, J. Singh, and C. Prakash, “Prediction of minimum fluidization velocity for fine tailings materials,” Powder Technology, vol. 196, pp. 263-271, 2009. J. M. Begovich and J. S. Watson, “Hydrodynamic Characteristics of Three Phase Fluidized bed,” Cambridge University Press, pp. 190-195, 1978. V. R. Dhanuka and J. B. Stepanek, “Gas, liquid hold up and pressure drop measurements in a three-phase fluidized bed,” Cambridge University Press, pp. 179-183, 1978. H. Abukhalifeh, M. E. Fayed, and R. Dhib, “Hydrodynamics of TBC with non-NEwtonian liquids: Liquid holdup,” Chemical Engineering and Processing, vol. 48, pp. 1222-1228, 2009. I. Hamdad, S. Hashemi, D. Rossi, and A. Macchi, “Oxygen transfer and hydrodynamics in three-phase inverse fluidized beds,” Chemical Engineering Science, vol. 62, pp. 7399-7405, 2007.

H. M. Jena, G. K. Roy, and B. C. Meikap, “Hydrodynamics of a gas-liquid-solid fluidized bed with hollow cylindrical particles,” Chemical Engineering and Processing, vol. 48, pp. 279-287, 2009. K. V. Ramesh, G. M. J. Raju, M. S. N. Murthy, and C. B. Sarma, “Wall-to-bed mass transfer in three-phase fluidized beds in the absence and presence of a composite promoter,” Chemical Engineering Journal vol. 152, pp. 207-211, 2009. K. I. Lee, S. M. Son, U. Y. Kim, Y. Kang, S. H. Kang, S. D. Kim, J. K. Lee, Y. C. Seo, and W. H. Kim, “Particle dispersion in viscous three-phase inverse fluidized beds,” Chemical Engineering Science, vol. 62, pp. 7060-7067, 2007. H. J. Subramani, M. B. M. Balaiyya, and L. R. Miranda, “Minimum fluidization velocity at elevated temperatures for Geldart’s group-B powders,” Experimental Thermal and Fluid Science, vol. 32, pp. 166-173, 2007.

414

International Journal of Bioscience, Biochemistry and Bioinformatics, Vol. 2, No. 6, November 2012 M. Thirumarimurugan has pursued his Ph.D. in heat transfer from Anna University Chennai, Chennai. He has 17 years of teaching experience and currently holds the post of associate professor in Department of Chemical Engineering, at Coimbatore Institute of Technology, Coimbatore. He has authored one book, 13 international journals, 3 national journals and 51 conference proceedings. He has guided 16 PG projects, 22 UG projects, 4 mini projects and he is currently guiding 5 Ph.D. research projects. He is a life member of IICh.E and a member in Indian Society for Technical Education.

D. Prabhakaran was awarded the Ph.D. degree in chemical engineering from Anna University Coimbatore. He has 20 years of teaching experience. He is currently the associate professor with Department of Chemical Engineering, Coimbatore Institute of Technology Coimbatore, Tamil Nadu. He has presented papers in more than 20 national and international conferences and guided a number of projects for M.Tech. and B.Tech. and also published more than 10 papers in national and international journals. He is a life member of Indian Institute of Chemical Engineers and Indian Society of Technical Education, and a member of Asia Pacific Chemical, Biological and Environmental Engineering Society (APCBEES).

415